Conor Grennan on Moving Beyond the “Search Engine Mindset”

MOLLY WOOD: Today, I’m talking with Conor Grennan, Chief AI Architect at the NYU Stern School of Business and founder of the training and consulting company AI Mindset. Grennan helps aspiring entrepreneurs, leaders, and professionals think about how to approach AI for business, which is kind of what we do here too. One of his core messages is that we need to adopt a new mindset and break free of an outdated search engine paradigm and embrace an iterative, conversational approach for the AI era. In our conversation, we discussed how leaders can seize the potential of AI, but more importantly, how they can shift their mindset for the new AI era. Here’s my conversation with Conor. 

MOLLY WOOD: Conor, thanks so much for coming on WorkLab.  

CONOR GRENNAN: Thanks, Molly. Great to be here. 

MOLLY WOOD: I feel like a lot of the guests we talk to had a moment, kind of a eureka moment, a breakthrough moment that made them go, okay, this is going to be a big deal. I should probably start a consulting company around it—at least in your case. [Laughter] Can you tell me what that moment was for you? 

CONOR GRENNAN: For me, it was a pretty stunning moment because my wife was calling from the other room, like, hey, have you seen this? And I was like, no, what is it? And she’s like, well, just open it up. So I opened it up and all I see is a search bar, right, and she’s like, Oh, it’s a large language model. And I’m like, I don’t know what that means. So she’s like, well, just ask it a question. And I’m like, well, like what? And she’s like, you know, like anything. And I was like, okay, yeah, totally, totally… so, what kind of thing? She’s like, just say anything. I’m like, right. And I just sort of dove in and I’m like, all right, explain string theory to me. And it happened so fast and so brilliantly, and for the next three hours, I just went down this rabbit hole. And then the very next thing I did was text my boss at NYU Stern School of Business—this is when I was dean of students—I just said, hey, we got to do something. This is going to change everything. And he was like, I totally agree. So I started to build out a framework that we could teach our students. That was my only interest. I didn’t need to start consulting or anything like that, but that’s how it started. 

MOLLY WOOD: So now you’re preparing students for the future, and also leaders and professionals through this consultancy. Are there commonalities in how you approach that? I mean, it’s kind of a rare time, right, where everybody’s at the starting line in some ways. 

CONOR GRENNAN: It’s a deep secret that I’m reluctant to unveil, but here we are, the sentence has already started, so…. Yeah, the commonalities are almost a complete overlap, even between industries. And I’ll tell you what I mean by that. So I used to go out and work with companies, trying to get use cases for our students and for our faculty and administration, everything else. And I would be going out and saying, Hey, are you using these tools? And they would say, No, not really. And I would say, well, they’re really useful. You should try them out. And they would ask me to come out and speak to C-suite–level people, kind of across industry. And I would do that and teach them this framework of how to use these tools. And they would say, Oh my gosh, this is amazing. But I would go to something like a healthcare organization and they would say, Hey, come in prepared with the top 10 use cases for healthcare, right? Because that’s the history of how we teach technology: Give us the use cases. It makes total sense. You know, I would look up use cases for healthcare, because what do I know about healthcare? And I would teach them how to use these tools. And then I would check back in—and, by the way, they were like, This is amazing. You’re the greatest teacher of all time. I’m like, that’s right, I am. Then I would check back in maybe a month later, I would be like, Hey, how’s it going? And they’re like, Oh, those use cases are amazing. They’ve transformed everything. And I’m like, but I didn’t know anything about it. Like, why are you using my use cases? And I started to realize that the use case model of teaching just wasn’t very effective. I started to realize that no matter who you are, in any industry, it actually all required a behavioral shift. It doesn’t matter if you’re in private equity or healthcare or operations or anything—everybody thinks that their industry is different. It’s not, it’s all the same because it’s how humans think. That’s my experience. 

MOLLY WOOD: Let me break that down a little bit. So you’re saying that, traditionally, the way you would approach teaching, either business leaders or students, a new skill is this pretty well understood use case model, but because this is so new and is fundamentally not necessarily about only applying technology, it’s really a mindset shift; that you’ve had to develop a new kind of framework. 

CONOR GRENNAN: Yeah, a hundred percent. I actually had to tear up this framework that was actually really working. That’s a very hard thing to do when you have a framework that’s working, because what I was seeing in organizations was that they thought they were transforming, but it just meant that they were using generative AI—large language models, whatever it was—and they were speeding up some processes, but it wasn’t transforming how they were doing business across every department. And I knew that it could. And I also knew that there was essentially no learning curve to this, right? So, traditionally, when you try to figure out a new tool there’s a learning curve, right? If you’re thinking about learning French or calculus or Excel or something like that, right, there’s a learning curve and then you learn it and then you’re fluent in that thing. But large language models, they don’t have a learning curve. People think they do, but they actually don’t. You just do it. It’s sort of like exercise, right? Like, we already know we eat less and exercise to get in better shape, but there’s nothing to learn. And so I started to realize like, oh, there’s not something I really need to teach. I think there’s this sense, even with a product that’s as, you know, I think as fantastic as something like Copilot, there’s a sense that if you kind of give Copilot to everybody, everybody’s just going to start using it for everything. But it’s just not true. It’s sort of like thinking that if we put a treadmill in every home in America we’re going to cure heart disease. It doesn’t, because the problem is not learning how to use the treadmill. The problem is changing our behavior. And it’s so different. It’s so unusual because it’s really a partner. Mustafa Suleyman, who’s CEO of Microsoft AI, had this phenomenal quote: What he’s saying is that it’s no longer that we have to learn the language of computers; they’re going to learn the language of us. And that’s going to change everything, I think. But in order to do that, we have to think a certain way. So I’ve spent a lot of time working on how to change that behavior—not just showing people a tool, but how do we change the behavior on it? 

MOLLY WOOD: And a lot of your leadership advice now boils down to, really, this “bring your whole self” concept, right? Like, you’re telling people, don’t set aside your interests, don’t set aside your passion and just throw some AI on something. Really explore how to bring that partnership in line with your passion. 

CONOR GRENNAN: Yeah, that may be my business school bias as well. So I’ve been at NYU Stern for, you know, 11 years or something like that. I got my degree from there, everything. So my bias is really watching, you know, markets go up and down, economies go up and down, different trends happening, and watching, understandably, MBA students and others, following those trends. And so now you hear people saying, Oh, I really want to get into AI. And my advice is, well, do you? I mean, is that actually your interest, or do you want to use AI to completely transform the thing that you’re actually passionate about? So I think that we think adoption is very steep right now, but I don’t think it is. And what that means is that there is a tremendous amount of opportunity on the table for everyone to transform not just their work, but really their organization. That’s where we are. When we think about, well, where is adoption right now? I know that Microsoft and LinkedIn did a study together around this, and they were grappling with this. It’s like, well, is adoption high or is it not? Because it looks like a lot of people are using it, but there’s still a tremendous amount of need. So how can those two things exist simultaneously? And it took me a long time to work out why that would be. But if I was in front of a room of 500 and I said, how many people are using Excel, or something like that, everybody might raise their hand. But if I said, well, how many people are using it 30 times a day to build, you know, multibillion-dollar models to sell one… you know. Three investment bankers will raise their hand and everybody will be like, no, I just use it once a week to do a grocery list or to do a personal budget. That’s generative AI, you know, it’s this sense of, yes, everybody has tried it. Yes, people use it—I’m air quoting here—everybody uses it a few times a week to write emails. That’s different from what this can do. And the market is so wide open right now. So I really encourage people to follow what they’re actually interested in. 

MOLLY WOOD: You just mentioned the idea that incorporating AI well can transform not just your individual productivity, but an entire organization. Can you say more about that? 

CONOR GRENNAN: Yeah, I think, you know, with other tools, I don’t know, CRMs or whatever they are, the way that these typically work is that you bring them into an organization, and it’s a fairly typical digital transformation. You say, not every department is going to use this customer relationship management tool, but the ones who do, you know, you have to use this new system, and we’re going to train you how to use it, and then we’re going to burn down the old system because everybody wants to use the old system…. 

MOLLY WOOD: And that transition will be super messy, and a lot of stuff will get lost, and you’ll all be mad. 

CONOR GRENNAN: And you’re going to have the people who are like, never! The old pencil and paper forever! You know, that kind of thing. But the people who adopt faster will end up doing their work faster. And eventually you’ll have everybody moving up this slope and you’ll have that done. But there’s very few tools that are so widespread that you can use them across any knowledge worker’s role in any different kind of company. And so what I think we have right now is we have some folks using it very, very well. And then others thinking, well, you know, we have to do this better. I work with a lot of senior leadership teams at very large organizations, and the reason why I always work with senior leadership first, they have to set the new benchmarks for the organization. They have to see, oh, well, if it does this—really show them what Copilot in the hands of all their people would look like. But that’s not about, let’s show you the top 10 uses for Copilot. Instead, you have to show them what it will look like to change the way they learn, execute, strategize, communicate—everything. And that’s applicable across any knowledge work. And then what they have to do at the senior leadership level, is you have to set benchmarks for what a new eight-hour day works like. Otherwise, you have some people using it. It throws off talent evaluation. Can you imagine, you have some people using this and some people not? It’s like, well, so you can get to work faster and you don’t realize one dude has a bike. It’s one of these things where you have to set new benchmarks for that. And that’s the interesting thing to me—how we infuse this across the organization. 

MOLLY WOOD: I just want to spend hours and hours on that question of talent evaluation. Like, we talk a lot about, how are you going to tailor your hiring and your recruiting and your training around this, but the evaluation part of it seems huge. You see somebody really pulling ahead in this race. You have to ask, how are they getting there? And why is the person next to them maybe not? 

CONOR GRENNAN: You know, what’s the performance-enhancing drug that they are using, so to speak on this, right? And it’s funny because that’s the part where, if you can get your entire organization leveled up, the productivity gains are vast. The way that I think about it is that right now we’re locked into this mindset of thinking, well, how do I use it in my department? Well, how are you using it? Well, let’s bring in somebody, you know, who and what department is using these kinds of tools. Well, who’s the person that’s using Copilot really well, let’s bring them in and they can share. The problem is if you were trying to convince me to do morning yoga, right? Like, you may know that morning yoga is good for me, but I’m not doing morning yoga just because you say so. It’s a behavioral shift, and it’s a problem because it’s a behavioral shift. So, in other words, if I brought in to an organization that only had candles, like, I brought in the flashlight, everybody’s going to start using it because that’s not a behavioral shift. That’s just, this is much easier. But the brain has trouble with this, because when you’re looking at something like Copilot, and especially the real, how is this going to augment any question I have for it, the issue is that that’s a hard thing to get to use normally, right? Because it’s sort of like, it looks like a search bar, essentially. And our brain has trouble with that. So what I really talk to leadership about is it’s going to, as you articulated, it’s going to throw off talent evaluation, because if some people are using the internet or some people are using, you know, cars and bikes and other people are walking, how are you supposed to know what your best people, augmented by these tools, can look like? So you have to change behavior and that’s hard, but I think there’s a way to do it. 

MOLLY WOOD: I can’t tell, just as a side note, if you’re super into exercise or super resistant to exercise…. 

CONOR GRENNAN: A little bit of both. Mostly resistant. Mostly I love the couch. I love the TV. [Laughter] 

MOLLY WOOD: So, one of the other things you said is that English majors will rule the AI era, and we are discovering these interesting things about how you can get more out of LLMs by being conversational, and that there are certain populations who are actually a little bit better at that. I think I read some research that said boomers actually sometimes get better results because they’re not as, kind of, transactional with LLMs. They’re like, well, I don’t know about that, maybe explain this a little more. Talk about that idea of, heaven help us, the resurgence of communication skills! Yay! [Laughter] 

CONOR GRENNAN: It’s amazing, right? I mean, talking about other universities that I talk to and that I commune with. A lot of them are saying, you can’t just come in and do computer science anymore. You have to have a humanities minor or something like that. We still definitely need technical people. But what I would say is that the history of technology has been that if you are a little more digitally savvy, say, you would be able to handle the next digital-savvy thing that comes along. And this is the exception. And why is that? Because it is a software that doesn’t really behave like a software. It behaves like a human. And the more that you can have a conversation with it—now that requires changing your brain and changing behavior, and again, that’s what I focus on with organizations, but that’s the critical part. So if you can do that, well, who’s very good at that? Well, it tends to be people who, you know, I call it English majors just to sort of be a little bit provocative, but it’s people who are pretty decent at communication, whether it’s writing or speaking or, you know, good readers, things like that. Because they understand what a good conversation flow is, they understand what it looks like to guide other people, and English majors and the like have an amazing advantage here, because if they can just crack through their sort of search bar brain, where you’re used to command, response, walk away, and instead have an iterative conversation, and they’re very well poised to do that. And this is where the behavioral shift comes in. Then they will all of a sudden have this companion that will bring them to the moon. And it’s really, really a powerful tool. 

MOLLY WOOD: Yeah, say a little more about the search engine mindset. You’ve talked about how this can be a barrier to this kind of behavior change that’s so necessary. 

CONOR GRENNAN: It was the thing that made me tear up my framework, which again, was extremely painful to do when you have a framework that works. But you’re like, but it’s not transforming. 

MOLLY WOOD: I know a fair number of teachers and professors—I know this pain that you’re talking about. You’re like, I have my curriculum. I’m good to go. [Laughter] 

CONOR GRENNAN: It was so good. [Laughter] But it’s the transformational power of this thing. And again, a lot of times people say, Well, we don’t want to start with transforming. Let’s start with small use cases. I say, no, no. Here’s the reason why not: because if our brain loves, you know, pattern prediction, automation, things like that, it can really steer us in the wrong way. So the reason I don’t teach through use cases is because then it assumes that, well, when this use case comes up, you pull it off a shelf and you use that use case, right? And that’s a problem because it really narrows how you think about it. So one brief analogy before I get to the search bar too, is that if you were working in, you know, pre-whatever, 1850, or whenever it was, and everybody had candles on their desks and then you gave them light bulbs. Yeah, that’s great, but a light bulb is only a use case for electricity. So it’s essentially saying, a light bulb is the same thing as all electricity, but this other factory next door, yeah, you have better light. The factory next door is using electricity for their payroll and for their trucks and to ship these things. And they’re using sewing machines to sew their clothes. They’re using it for everything. And you can see that difference. So factory number one is like, this is so much better than it was. The search bar problem, it kind of gets to that same idea, which is what’s problematic when you see a large language model? The problem is that it looks like a search engine. Now people say, I know I kind of treat it like that, but I don’t anymore. Yes, you do, because I do too and I’m on this stuff all the time. And why is that? Because our brains have these things called neural pathways, which automate things. So if you think about all the things you automate, you don’t even think about it, but like, when you see a baby, how do you not accidentally talk to it like a college professor? Because your brain has automated that, it’s the muscle memory. It’s why, you know, when you’ve been driving the same way to work for 20 years, you instinctively make the left out of your driveway because your brain has automated all these things. So that’s good. It frees up your free prefrontal cortex for thinking, all that. It’s bad if you’re trying to break a habit, and we are so used to the search engine from back in whatever it was, Netscape, or the original search engines, right? I mean, this is what we’ve been doing forever. We see that bar, we type in a question, we get a good answer because algorithms push the good answers, but then we walk away. To get people out of this, I say, well, imagine if you’re planning a trip to Costa Rica and at one table here, you have a laptop and Bing is open and you say, Hey, how should I, what should I do with my family in Costa Rica? It’ll give you a great response, right? But if door number two led you to an hourlong conversation with the head of the Costa Rican tourism board, and he would say, well, so what do you want to do? What are your kids into? Do you have any, well, how long do you actually have? Well, you know what we could do? We could switch it around instead of doing that, and, you know, I know we say this, but this new thing… that’s a conversation. And that’s what large language models are. They’re the head of the Costa Rican tourism board. But it’s hard because they don’t look like it. They look like a search bar and your brain has a lot of problems with that. 

MOLLY WOOD: How do you even start there? Like, it’s the tendency of organizations, right? I am just now picturing an organization, thinking about adopting a new tool, thinking, I will probably roll this out to my technical teams first. I’ll pilot it in some way. And you’re just going to have to sit down and have a conversation with them. I mean, that just—no wonder you’ve had to throw out your curriculum. 

CONOR GRENNAN: A hundred percent. Because, again, the tricky thing is that you see improvement, but you don’t see transformation. It’s not easier to do slight improvement then complete transformation, if you think about it, because it’s not like normal digital technology where there’s easy use cases and advanced use cases, right? It’s just another brain for you. So that’s the strange thing, is that it’s not like, well, I’ve been using it for this long, so I’m this good. So what I talk to businesses about, and senior leadership teams about, is that you have to get educated. You have to find somebody that you trust that’s going to have a way of changing behavior. If you’re just bringing people in and saying, Hey, we’re going to show you how to use Copilot, it’s not going to work. So you have to have a formal, structured education from somebody that you trust, an organization that you trust. And the way that I tend to approach that is, first of all, making sure senior leaderships understand it, focusing completely on behavior because you have to do that, and then following up and making sure that they’re actually changing how they do this. It’s difficult, because when you have something as powerful as Copilot that can change the way everybody works and you’re putting it at people’s desks, they have everything they need. They don’t need to learn anything else. And that’s what kind of breaks my heart sometimes. 

MOLLY WOOD: I wonder when you think about investment, like implementing these tools as an investment of money, you gave an example and said, look, if you were going to distill and summarize, if somebody asked you to distill and summarize something all by yourself, it would take hours. If you asked an LLM to do it for you, it would take 30 minutes, but you have to invest that 30 minutes. 

CONOR GRENNAN: Right. Yeah. 

MOLLY WOOD: And so how does that translate to rolling out a tool like this to an organization and saying, we will give you the time and space to learn and understand this, as opposed to, we expect those performance metrics to change immediately. 

CONOR GRENNAN: So the way that I see it, personally, and I guess every organization is different because every culture is different. I’ll just say that the way that I’ve seen it work and the way that I advise, and coming from a business school is helpful because, you know, I’ve seen a lot of different industries over the last 10 or 11 years that I’ve been there. But the way that I think it works is that first of all, this is an incredible employee engagement tool, and you have to think of it like that. This is why I often say this should live in the HR, talent world, and I’m not saying they should make the decisions on the tools or anything like that. But you do have one department that understands talent and understands what motivates talent. Now, nobody takes a CRM home with them and makes their life better or uses it to help them become a better spouse or father or anything like that. But with LLMs, you can. It’s an amazing upskilling tool. So anybody in your organization, I’m guessing, is going to leap at the opportunity to actually learn this because they see other people using this all the time. Also, this is going to help them plant tulips, and it’s going to become their IT help desk, and it’s going to become, like, helping them write Mother’s Day cards, and everything else at home. Because no company can say, Hey, this is going to help you become more productive for us, and then you should use it. What do they care? I mean, they care, I know, but you know, in a big organization, it’s hard for people to care about the fractional increase in the bottom line. So instead you have to say, we’re going to upskill you for whatever your next role is going to be. We’re going to upskill you so that you can use this tool when you go home. And also, maybe most importantly, this is going to get rid of so many non-value and menial tasks, or at least condense them. Let’s say we’re going to save 45 minutes a day—this is where we get into the time management piece, which I think has to go hand in hand—what are you going to do with that 45 minutes? Because it is easy to keep fighting fires and dive back into your email. But if you said to your people, listen, we’re gonna give you 45 minutes back. And with that, take that and do something you’re passionate about. Try a side project, work with a team on some—give them back that kind of thing. And we’ve seen that kind of thing work before, where you allow people to be more themselves. When I work with sales teams, they say, we just want to be more ourselves. You can imagine this with doctors, with salespeople, with all these folks who are just, You hired me to do this, but maybe 60 percent of my role is a lot of paperwork. What if we got rid of that and allowed that person to be them? There’s just so many opportunities here. 

MOLLY WOOD: I would like to see a world where that is the management approach. Where they say, We saved this productivity. We got this 45 minutes back. Because I think what you’re saying is it has to happen in partnership with your employees. They have to see a benefit, not just to the timesheet, but to themselves. 

CONOR GRENNAN: I think so. 

MOLLY WOOD: You last year told Business Insider that you use AI 70 times a day. Can you tell us about some of those unexpected examples? Inspirations? Ways that you’ve used AI in your work or personal life. 

CONOR GRENNAN: I use it so often, because it’s just constantly open on my machine. So, I’ll just think about the last couple of days because it’s, you know, it’s just sort of startling. So my family is planning a trip. It’s a lot of different flights because we’re going to Southeast Asia. And I went in and just copied all the emails and just put them into a large language model and said, Hey, come up with an itinerary, figure out all the cancellation policies, all that sort of stuff. Now people are like, I know we use it for travel. I’m not talking about using it for travel. I’m talking about using it for the little frictions. Otherwise, what would I have done? I would have created a spreadsheet, I would have gone into the email, looked up the stuff and typed it all out. Another thing that I do constantly is when something on my laptop isn’t—or I can’t get it to work, I’m sure it is—working, I just take a photo and I upload the photo. I’m like, Hey, this is what I’m seeing, what am I doing wrong? Oh, you’re on this, you know, machine, sort of like you’re on this PC. Okay, go over here, go to the Windows icon and click on this and then click on settings and click on advanced settings and click on…. It’s a constant IT help desk for me. It just depends on what you have to do. That’s the exciting part to me. 

MOLLY WOOD: Is there anything we haven’t already talked about with respect to AI opportunities, and maybe challenges, that you think leaders might be overlooking? 

CONOR GRENNAN: I just believe, I suppose, that people have to not underestimate how difficult this is. It’s a very, very simple technology in that you don’t need to know anything, but it is also difficult to get people to change their behavior. It’s much more akin to a cultural transformation or something like that, rather than, hey, we’re going to get everybody using this new tool. Don’t underestimate that. Don’t just take people who know how to use it and think they’re going to transform everybody else. Really understand that you need some kind of education, you need some kind of framework, you need something to guide you through it. 

MOLLY WOOD: Thank you again to Conor Grennan, Chief AI Architect at the NYU Stern School of Business. I appreciate the time. 

CONOR GRENNAN: Thank you. 

MOLLY WOOD: And that is a wrap for this season of WorkLab. Thank you for joining us. We’ll be back next season with more insights into this fast-moving landscape. If you’ve got a question or a comment, please drop us an email at worklab@microsoft.com, and check out Microsoft’s Work Trend Indexes and the WorkLab digital publication, where you’ll find all our episodes along with thoughtful stories that explore how business leaders are thriving in today’s new world of work. You can find all of it at microsoft.com/worklab. As for this podcast, please, if you don’t mind, rate us, review us, and follow us wherever you listen. It helps us out a ton. The WorkLab podcast is a place for experts to share their insights and opinions. As students of the future of work, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own, and they may not necessarily reflect Microsoft’s own research or positions. WorkLab is produced by Microsoft with Godfrey Dadich Partners and Reasonable Volume. I’m your host, Molly Wood. Sharon Kallander and Matthew Duncan produced this podcast. Jessica Voelker is the WorkLab editor.